Serijag

Hello, I have oobabooga setup in docker, it's not a CA template. I have it working when I only have one GPU running in Unraid but whenever I add a second GPU it tries to use both of them. I have the NVIDIA_VISIBLE_DEVICES set to the GPU ID from the Nivida plugin and nvidia-smi still shows both GPUs. I have tested both GPU IDs on the oobabooga container and it always shows both. On other container (tdarr, ollama, code-project) the NVIDIA_VISIBLE_DEVICES setting works to limit the GPU. Does the oobabooga container not support it? I thought it was a docker thing not container specific though. Here is the link to the docker hub and github for oobabooga docker I'm using. Attached are setup screenshots as well. https://github.com/Atinoda/text-generation-webui-docker https://hub.docker.com/r/atinoda/text-generation-webui/ Thanks!

It was totally fine from last night until this morning and it JUST happened within the last 30 minutes. Updated log files attached. darthvader-diagnostics-20250218-1001.zip

Yes, they were just taken a few hours ago and I have not restarted the server since the issue has last happened.

Hello, I upgraded to 7.0 from 6.12.13 and everything seemed to go okay except after a successful boot I will lose access to the webUI after a few hours. Tried looking through the syslog but wasn't able to find anything in the time from when it was working to when I can no longer access the webUI. SSH works fine and VMs and dockers all seem to be working normally as well. I can reboot the server from SSH and then webUI access will return for a few hours. The login page will still appear as normal but the broswer will either just spin or the UNRAID logo will appear but the rest of the page is just blank. Screenshot and logs are attached. Thank you! darthvader-diagnostics-20250217-1738.zip

Thanks, I ended up figuring it out pretty much following these steps. Firewall\router rules weren't working correctly after a modem reboot and had to readd the port forwarding rules for NPM. Never had that before so hopefully it was just a fluke. Thanks

I'm having this same issue did you ever figure out what was going on?

No I had to rebuild one of the VMs that I couldn't get to boot at all. I have 5 other VMs that all show that same message, some with GPU passthrough and some without. Even the new VM shows that message. Did you make any hardware changes? I only got that message after switching from AMD cpu\mobo to Intel.

Upgraded CPU and motherboard this weekend and now all VMs showing weird message after starting the VM. Most boot and don't seem to have issues but I didn't have rebuild one. Curious of the cause and if there is a fix, something to worry about, or just a little message. I did switch and an AMD configuration to Intel. Nothing is passed through on the VMs, no GPUs or storage. This does happen on both Windows and Linux VMs as well. Storage is mostly NVMe drives (not passed through, just a cache drive) and an SSD. Diagnostics and screenshot of the message are attached. darthvader-diagnostics-20230306-1323.zip

Ahhhh interesting I'll take a look at that. Thanks very much.

Never had any issues with server shutting down until upgrading past 6.11.1 and now the server will power off unexpectedly after about a day. No issues I can tell other than the server powering off. I have reverted back to 6.11.1 to see if it continues and was just unforunte timing or something else. Attached are logsdiagnostics-20221120-1430.zip